• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 1137 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM3478: +3.3V to +48.5V @ 50mA : Most efficient solution ?

jerome d.
Expert 1325 points
Part Number: LM3478
Other Parts Discussed in Thread: TPS7A4101, TPS61040, , LM3488, LM5155

Dear All
I'm designing a boost power supply, my requirements are +3.3V input (range from 3.2 to 3.4V), output voltage should be +48.5V, output current should be from 10 to 50mA.
I found different parts that should do the job, and was wondering if I could read some benefits/drawbacks using those different parts :
I'm designing this power supply in a portable device, efficiency and low Iq are quite important.

The final +48V should be low ripple, as it is dedicated to power a microphone phantom, my plan is to use a low noise LDO to convert the DC/DC switch boost 48.5V to a clean 48V output.
I plan to use a TPS7A4101 for this final stage.

What I found until now regarding the +3.3 to +48.5V :

1) TPS61040 with additional diodes to reach higher voltage than 28V : www.ti.com/.../slua272.pdf
2) LM3478 or LM3488

When simulating the LM3478 or LM3488 using Webench designer, the circuit seems to work fine until 48V, once I setup the output voltage to 48.2V the BOM cost hugely increases.

What would be the best solution for this design requirement ?
Is there another circuit I missed ?

Any help would be greatly appreciated.
Thanks

  • 0
    TI__Mastermind 21720 points
    Hi Jerome,

    The LM5155 might be a better fit for the application assuming that you operate in DCM all the time. Operating in DCM will allow you select many different parts that would otherwise not be able to do the application because of the duty cycle limit.

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Expert 1325 points

    Thanks Garrett for your suggested device.
    Unfortunately I cannot simulate this device using Webench, It always results in "Calculated duty cycle exceeds limit" error.
    I setup as follow :
    Vin from 3.5 to 9V (although my real input voltage is 3.3V)
    Vout = 48.5V
    Iout = 0.05A
    Any idea why ?
    Before updating my schematics with that part, I'd like to have a rough idea of efficiency at Vin=3.3V Vout=+48.5V Iout=50mA, as well as a rough idea of the ripple voltage at the output.
    Thanks
    Jerome

  • 0 in reply to jerome d.
    Expert 1325 points

    Garrett
    I'm facing some difficulties to calculate the external components values according to my requirements.
    As far as I understand, the inductor should be a low as possible to operate in DCM mode.
    Targeting at least 70% efficiency at light load (10mA) requires a switching frequency as low as possible, something like 220KHz (hopefully this will not result in harmonics in the audio band)

    At 220Kz switching frequency, the recommended inductor is about 10uH, selecting those value in the XLS calculation sheet results in a recommended external slope compensation resistor Rsl of 3K, with Rs = 8mR, while the design guidelines mentions Rsl should be less than 1K.

    Should I follow the 3K Rsl resistor recommended by the design tool, or should I definitely limit to 1K ?
    Limiting to 1K results in peak current inductor limit to about 9A, which is quite important.

    Any help would be appreciated.
    Thanks

  • 0 in reply to jerome d.
    TI__Mastermind 21720 points
    Hi Jared,

    This article should help you understand the design requirements for a Boost that always operates in DCM.
    www.ti.com/.../Topic_3_Lynch.pdf

    WEBENCH will not design the application because it is assumed that the regulator is going to operate in CCM. In CCM the LM5155 will run into the maximum duty cycle limit which is why DCM is required for a large step-ratio. To be in DCM the inductance should be small enough to ensure that the current is discontinuous when the input voltage is at the minimum and the output current is a the maximum.

    Please let me know if you have any questions.

    Thanks,

    Garrett